Aardvark DailyNew Zealand's longest-running online daily news and commentary publication, now in its 25th year. The opinion pieces presented here are not purported to be fact but reasonable effort is made to ensure accuracy.Content copyright © 1995 - 2019 to Bruce Simpson (aka Aardvark), the logo was kindly created for Aardvark Daily by the folks at aardvark.co.uk |
Please visit the sponsor! |
My gob was well and truly smacked this week.
NVIDIA announced their latest and greatest AI processor chip and it has specifications that simply blow my mind.
This thing has... wait for it... 208 billion transistors in it.
Let me write that again: 208,000,000,000 transistors.
Yep, every single processor has 26 times as many transistors as there are people on the planet right now.
Incredible!
To further prove how fast this semiconductor technology is progressing, it also sets new records for efficiency, using significantly less power than previous processors.
And here's me, still recalling how I marveled for weeks over the very first transistor I ever bought (an AC126 PNP germainium device from memory) -- now they're putting 208 billion of the damned things on a single die!
Moore's Law for the win!
As I cruise on towards the end of the rocky road that is life, I can't help but wonder where this sort of tech will be in another 50 years.
One thing is for sure, AI is changing the world and creating as many risks as it does benefits. Processors such as the new one from NVIDIA will only hasten the adoption of AI, for better or worse.
Some of the more sinister aspects of AI revolve around its ability to interpret huge chunks of data -- the sort of data that can be collected covertly by things such as smartphones and other devices.
For example, in the recent consultation here in Tokoroa over the repurposing of a reserve into a housing estate, the consultants hired by the council claim that they were able to ascertain the level of use that reserve got by "pinging" people's cellphones.
Really?
How did they do that?
To be honest, I think they're BSing about this -- but that's what consultants tend to do sometimes when they're asked to produce data to support an agenda.
The bigger problem is that with large AI systems it becomes possible for governments or private industry to do an awful lot of tracking and then analyse insane amounts of information to spot things that may be politically or commercially valuable.
Apparently this is the sort of thing that is happening in China right now where massive amounts of data is collected on each citizen and stored for analysis by AI. The goal here is allegedly to identify threats before they become active.
It's only a matter of time before all governments around the world engage in this Minority Report behaviour thanks to the power of AI and our propensity to carry our phones everywhere and pour our lives out onto social media.
Certainly the future is going to be a very different place to the past but we, as a people, will certainly have to be increasingly vigilant and intolerant of the rising level of surveilance and intrusion by "the powers that be".
Remember:
War is peace, freedom is slavery, ignorance is strength"
or so our overlords would have us believe.
Whilst on the subject of AI, I see that YouTube is about to require all creators to clearly label AI-generated content as such, where there is a potential to mislead people. Whilst this might appear to be a great idea, it also creates a potential problem for people like me who often use good, old-fashioned VFX in their videos.
When you see a police car sign-written with "Drone Enforcement Team" you can be pretty sure that it was just me using some compositing software and hard work to create an illusion. This is NOT AI however, I'm pretty sure that some troll will report such videos as being AI and YouTube being YouTube, those videos will get me in trouble.
In fact, creators are in a real pickle here because the creative use of VFX is quite essential to some forms of video creation but labeling this footage as containing "synthetic" content will drop it into the same bucket as those inane AI-generated spam videos with computer-generated narration of a Wikipedia page. Such videos are prolific on the platform right now and are nothing but clickbait crap. The "synthetic content" tag could be the death of those creative channels that use VFX and AI for good.
I also have to wonder whether all those TV shows and movies that will increasingly use AI-generated content will also be clearly labeled as such when uploaded to YouTube. I bet you big money that they'll get a pass on that one.
Carpe Diem folks!
Please visit the sponsor! |
Beware The Alternative Energy Scammers
The Great "Run Your Car On Water" Scam